A Comparative Study of Machine Learning Models for Breast Cancer Diagnosis Using the Wisconsin Diagnostic Dataset

Accurate and interpretable diagnostic models are vital for early detection of breast cancer. This paper presents a comparative analysis of four classifiers Logistic Regression (LogReg), Random Forest (RF), Gradient Boosting (GBRT), and a compact Multilayer Perceptron (MLP)on the Breast Cancer Wisconsin (Diagnostic) dataset. Classical models were optimised via grid search and stratified 5-fold cross-validation (CV), while the MLP used validation-based early stopping. On the heldout test set, Logistic Regression achieved the strongest overall balance of discrimination and calibration (AUC 0.996, AP 0.998), with competitive F1/accuracy. A computational analysis showed that Logistic Regression trains orders of magnitude faster than ensembles and the MLP, and still delivers near real-time inference; GBRT offered slightly faster per-sample inference but at a substantially higher training cost. These results reinforce that transparent, well-calibrated linear models can match ensemble and neural baselines on small structured clinical data, supporting their suitability for time-sensitive decision support.